Algorithm for predicting death in newborns with respiratory pathology and perinatal damage to the central nervous system on artificial ventilation

Abstract

The objective was to predict the fatal outcome of the disease in newborns on artificial lung ventilation by means of an intelligent analysis of the immunological database.Materials and methods. The retrospective clinical study included 108 mature newborns. Upon admission to the intensive care unit, on the 3rdday and at the end of the disease, the plasma concentrations of IL-1β, IL-6, IL-8, TNF-α, G-CSF, s-Fas, FGF, NO were determined by ELISA; the relative content of CD3+CD19–, CD3–CD19+, CD3+CD4+, CD3+CD8+, CD69+, CD71+, CD95+, HLA-DR+, CD34+; CD14+, CD3–CD56+ by immunophenotyping; relative content of lymphocytes with expression of AnnexinV-FITC+PI–, AnnexinV-FITC+PI+. By the method of decision trees, the rule of predicting death was formulated.Results. The patient is predicted fatal outcome if, upon admission to intensive care, he has the relative content of lymphocytes with expression of AnnexinV-FITC+PI+ ≥ 0.95 % and plasma concentration of G-CSF ≤ 1.46 pg\ml or G-CSF ≥ 1.46 pg\ml and AnnexinV-FITC+PI+ ≥ 4.75 %(specificity 98.68 %; sensitivity 96.97 %; accuracy 98.68 %).Conclusion. In newborns with respiratory pathology and perinatal involvement of the central nervous system on artificial ventilation, death is determined by the high activity of T-lymphocyte apoptosis mediated by the low plasma concentration of granulocyte colony stimulating factor.